Actually, I was impressed by the suggestion of using a clear day as a standard from someone who lives in Seattle.
It seems to be an excellent idea. Since your artificial lights are probably less intense than sunlight, if you can devise a way to block a known portion of the light, you can obtain a two-point calibration curve to refine your measurements. Obviously, great precision is not required, but we have no idea how non-linear the response may be or whether the response is saturated in direct sunlight. I used to use "neutral density" filters, which can be made from particle suspensions or even solid materials with holes that represent a known percentage of the total area. I would suggest an optical density of 1.0 (i.e., something that blocks 90% of the light) for obtaining your second calibration point. If you have contacts in a meteorology dept as suggested by Philba, they may have filters you could borrow. John